1020 
around the coasts bordering the North Atlantic Ocean. 
This work established that ablation of a glacier is 
controlled by two factors: direct radiation, which 
predominates in dry continental climates; and a com- 
bination of convection and condensation, which pre- 
dominates in moist maritime climates. It established 
further that temperature is the dominant factor in 
the glacier regimen; glaciers are more responsive to 
changes in temperature than to changes in precipitation. 
Correlation with Climatic Changes Determined Inde- 
pendently. Not until recently was the probability of a 
contemporary widespread temperature increase during 
the last 100 years established by actual compilation of 
weather records [10, 11], with the change affecting both 
polar hemispheres. Knowledge of this temperature in- 
crease was refined and enlarged through the subsequent 
appearance of several more compilations and interpre- 
tations of weather records (summarized and cited by 
Ahlmann [4, pp. 169-181)). 
The modern temperature increase has been concomi- 
tant with a general shrinkage m the volumes of glaciers 
in both hemispheres [8, pp. 67-68; 15, p. 231]. It has 
been concomitant also with a reduction in the extent 
and thickness of floating ice in the Arctic Sea [4, p. 187] 
and with an apparent rise of sea level [9, 13], best ex- 
plained as the result of an increment of water derived 
from the melting of glacier ice. 
There appears, then, to be an obvious correlation 
among these observations: (1) secular increase of mean 
temperature, (2) shrinkage of glaciers, (8) shrinkage of 
sea ice, and (4) apparent rise of sea level generally 
within the span of the last 100 years. 
This correlation seems to establish the relation be- 
tween the fluctuations of climate, although data both 
more extensive and more precise would be desirable. 
Ever since the pioneer announcement by Venetz, re- 
search workers have assumed that pronounced former 
variations in glaciers reflected changes in climate, and 
on this basis have inferred major climatic variations 
from the abundant geologic evidence of variations in 
glaciers. When ‘the recent apparent correspondence be- 
tween glaciers and climate has become more firmly es- 
tablished, the resulting quantitative data should prove 
useful in interpreting the climatic variations of earlier 
times. 
Changes in Glaciers within Historic Time 
As is true of most geologic phenomena, both the 
abundance and the quality of the evidence with which 
we have to work diminish conspicuously as we go back- 
ward in timé. Therefore, in a review of the climatic 
evidence furnished by glaciers it is best to begin with 
the present and work back into the past. 
Changes within the Last 100 Years. Abundant evi- 
dence supports the statement, made above, that glaciers 
have been generally shrinking during the last century. 
The evidence is derived from observations made in 
western North America (including Alaska), Greenland, 
Iceland, Spitsbergen, Scandinavia, the Alps, East 
Africa, South America, New Zealand, and the Ant- 
arctic Continent [8, 4, 15, 20]. Correlated with this are 
CLIMATOLOGY 
a contemporaneous shrinkage of lakes and a group of 
ecologic and oceanographic changes (summarized in [4]). 
As Matthes [15] pointed out, the fact that glacier 
shrinkage has been occurring simultaneously in both 
polar hemispheres bears on existing hypotheses of the 
cause of climatic changes. This fact disfavors those 
“astronomic hypotheses” which demand, at least to 
some degree, temporal offsets of climatic effects be- 
tween the polar hemispheres. 
Changes since the Beginning of the Christian Era. 
Research in the historic period before the beginning of 
systematic glacier observations has consisted of extend- 
ing the data on glacier changes backward in time 
through the study of official records made for other 
purposes. This was attempted first, and with consider- 
able success, by Venetz [22]. Another example is a study 
of the archives of the town of Chamonix in the French 
Alps. In this study Rabot [18] succeeded in extending 
the record back to the year 1850 and demonstrated 
further that the present-day shrinkage commenced, in 
that district, about the middle of the nineteenth cen- 
tury, following a period of some 250 years during which 
the glaciers were relatively expanded. A more compre- 
hensive reconstruction, covering a much wider area 
although it extended back only to the year 1700, was 
attempted by Briickner [5]. 
Similar research into the Iceland record by Thorarins- 
son [21] confirmed and extended these conclusions. 
Thorarinsson showed that glaciers on that island were 
far less extensive from the tenth century to the thir- 
teenth than during the period since the beginning of the 
fourteenth century, and established changes of lesser 
degree within the latter period. This reconstruction is 
supported by various historical data not directly con- 
nected with glaciers (e.g., [12, 23]). 
Most of the first millenium of the Christian era is 
believed to have been relatively warm, but the evi- 
dence is derived from the fluctuations of lakes, from tree 
rings, and from other nonglacial data. Apparently noth- 
ing is on record as to the condition of glaciers during this 
time. A record may be present in the stratigraphy of 
glacial deposits, but if so, it does not seem to have been 
recognized. 
Variations between the Wisconsin Maximum and the 
Beginning of the Christian Era 
Still farther back in time the glacial evidence consists, 
not of observations on the glaciers themselves, but of 
stratigraphic and morphologic features. Many of these 
features are moraines built along the glacier margins 
and abandoned during shrinkage; from these, former 
glacier dimensions—extent or thickness or both—can 
be approximated. Research on this ancient period lies 
in the field of glacial geology rather than in that of 
glaciology. 
The Wisconsin maximum is believed to represent the 
latest of a series of four major glacial expansions on a 
world-wide scale. It is generally assumed to date from 
about 60,000 years ago. The rough measurements avail- 
able indicate that at that time glacier ice, mostly in 
the form of great ice sheets, covered about 27 per cent 
